The above study examines the different factors that may influence education and there relationship to education attainment. To do the above hypothesis testing and regression analysis was used from data collected via online system. Factors studied include relationship between education and age, relationship between education and gender, relationship between marital status and education and influence of parental education on their children education. Gender and education attainment are negatively correlated with no significant linear relationship. Similarly parental education is important in influencing schooling attainment of any child. Nevertheless education attainment from both parents will not influence or make one stay in school more. Age on the other hand is an important factor when studying education attainment as per years one remained in schooling. It is negatively correlated though the relationship is also not linear.Respondent marital status and highest year of school completed are negatively correlated e.g. those widowed had the least mean years in school while those not married had the highest mean years in school and a majority of those married have attained graduate level while very few widows have been to graduate level.
Various factors have been put forward to explain the difference in education through years one decides to stay in school and their highest level of schooling. Comprehensive study of the factors is imperative so as to understand and comprehend how to improve education within any given region. Past studies have been undertaken but unique to their regions in trying to explain factors influencing education and their relationship to schooling attainment. Within the United States no country wide comprehensive studies have been done to try and explore the factors that determine education attainment. It’s from these bearing that this study is necessary.
This study aims at observing relationships between certain factors that are thought to influence education attainment and state their influence on education if any. To achieve the above goal it is necessary to meet answer the following questions which serve as the objectives of the study.
What is the relationship, if any between education and gender? Discuss any differences that may exist and describe the characteristics of the sample.
What is the relationship, if any, between parental education and the education of the respondent? If a relationship exists, which parent has the strongest effect on the educational level of the respondent?
Is there a linear relationship between age and education, and if so, how strong is that relationship? Is it possible to predict educational level based on age? If so, what limitations exist for the developed method?
What is the relationship of marital status on education? Do singles or married persons tend to be more highly educated?
Given the research questions appropriate hypothesis will have to be formulated and relevant test done to realise the objectives.
A web based tool was developed and hosted on the university servers with a link provided from the universities websites and the student online store that can be accessed by any one. The above was necessary to minimise on cost of creation of data collection tools in terms of printing and data collection process. The student online store is accessed by any one and has a daily traffic of over 1,000 visits. Any one visiting the site was considered a prospective interviewee and so the request to assist in carrying out the survey popped out as a link to direct one to the online survey questionnaire.
A total of 1417 respondents took part in the study. These were spread throughout the states of the country. The spread was necessary to ensure the results could be generalised and used to inform on a wider group with similar characteristics.
For the study the following information was collected from the respondents so as to achieve objectives of the study: Age of respondent, Respondent’s highest degree, highest year of school completed, Respondent’s sex, highest year school completed, respondent’s mother highest year school completed, respondent’s father respondent’s marital status.
The variables used for the study were of three measurement types:
Ordinal: used to show variables whose possible responses are ordered in a given way, variable includes respondent’s highest degree.
Nominal: used to define groups within response data sets no given order is followed and interval between numbers has no meaning, variables include respondent’s sex and respondent’s marital status.
Scale/interval: These measurements include respondent’s age, respondents highest year of school completed,respondent’s mother highest year school completed andrespondent’s father respondent’s marital status.
After data collection, data was processed and analysed using appropriate data analysis tools and statistical tests. To answer the research questions a null and alternative hypothesis will be postulated and two tail ttest of used to check if the relationship witnessed are linear of not. Rejection or non-rejection of hypothesis will depend on postulation of null hypothesis. To effectively carry out the test a confidence interval of 95% will be considered at an alpha level of 0.05.
Correlation and regression analysis will be necessary if the answers are to be achieved at objectively. Regression analysis will provide the R squared value that will tell how much of the variation noted in the dependent variable can be explained by the independent variable. It will also provide the multiple R which is the correlation coefficient that will provide the strength and direction of relation between the two study variables. A t test will be used to determine whether there exists a relationship between the two study variables in the different questions. A linear regression model will be used to model the existing relationship between the variables.
Relationship between education and gender
In order to understand the relationship between education and gender it will be necessary to study how correlation between the two variables, how much variation in one variable can be explained by another, appropriate distributions and test for the existence of the stated through correlation coefficients.
The relationship between education and gender can be explained using Pearson coefficient from regression analysis. The Pearson coefficient provides for strength and direction of relationship between two study variables. Table 1a below shows results of covariance analysis that will assist in provision of the Pearson coefficient so that establishment of the relationship between the study variables is realised.
More females’respondents attend school to the bachelor degrees level than men though more men than women proceed to postgraduate level. In the study Gender relations in primary and secondary education in Flanders a similar trend can be witnessed between bachelor and graduate level where more girls finish their bachelor’s degrees than male but more males proceed for undergraduate studies.
To test and ensure that the relationship between gender and education exists and is not as a result of chance it’s necessary to carry out a t test on the study variables. We shall assume the null hypothesis that there exist a relationship between gender and education and the alternative hypothesis shall negate the null hypothesis by stating that there exists no relationship between the two variables. Meta-analysis studies done by USAIDs office of women development by EQUATE project do not portray any similarities with the above findings other than more males proceed to postgraduate levels than women. There studies across ASIA and Africa where they fund projects shows that very few females even go past high school as most are resigned to early marriages and education preference is always directed to their male counterparts.
The t test statistic from the test is equal to 55.65 and its corresponding p value equals 0.00. given thatp value < 0.05 we reject the null hypothesis that there exists a relationship between gender and education level. Thus we can state that the 8% variation in education that could be explained by variation in gender may be as a result of chance.
In order to understand the relationship between respondent’s education and their parents education it will be necessary to study correlation between the two variables, check how much variation in one variable (dependent) can be explained by the other variable (independent), appropriate distributions and test for the existence of the relationship identified and stated through correlation coefficients.
The relationship between respondent education and their parent’s education can be explained using Spearman’s rho from regression analysis. The Spearman’s rho provides for strength and direction of relationship between two study variables. Table 2a below shows results of covariance analysis that will assist in provision of the Spearman’s rho so that establishment of the relationship between the study variables is realised.
A majority (over 50%) of those respondents who have undertaken postgraduate studies are married. Second to the married are the never married at about 25%, devised at slightly less than 20%, those separated and widowed jointly account for about 5% of total respondents who have undergone postgraduate studies. Almost a similar trend is maintained within the other levels. This implies that married men are more highly educated than those never married.
Gender and schooling attainment are negatively correlated and although no significant linear relationship exists between the two ages and education, a change in one will result in change in a different direction in the other. This variation is evident in the different schooling attainments between males and females. From the study more girls attend school to undergraduate degree level than boys although more males proceed to postgraduate studies than females. These findings are contrary to most studies done on relationship between gender and education. Past studied from other areas indicate more males tend to go to school and stay in school longer than females. The above findings could indicate an improved education system to cater for the female child although it could also spell neglect on the boy child during the early stages of schooling. Nevertheless the differences noticed will have to be investigated further to ascertain there significance.
Parental education is important in influencing schooling attainment of any child. Nevertheless education attainment from both parents will not influence or make one stay in school more. Parents of same gender to their children e.g. father and son does influence education attainment of their children. A similar influence would be provided by a mother to her daughter. The above findings do not support a mother’s education attainment as an influencing factor to her son’s education attainment or a father to her daughters. There is no linear relationship between parent education attainment and respondent’s education attainment. To establish the type of relationship existing from the provided correlation coefficient it is necessary to carry out further analysis of the data is necessary.
Age and highest level of schooling are correlated as can be seen from, the stated correlation coefficient of -0.163 nevertheless there exist no linear relationship between the variables. The mean age of those with highest level of schooling as less than high school their mean age is 52.86 while that of those with a bachelor’s degree is 43.21 implying that age is negatively correlated to education attainment which is in line with our correlation coefficient of -0.163. Nevertheless the above doesn’t hold for those doing post graduate studies as there average age is 49.18 implying the relationship isn’t linear. Tests done on the data confirmed the above results. 3% of Variation in highest schooling years among the respondents can be explained by variation in respondent’s age. Age is thus an important factor when studying education attainment as per years one remained in schooling.
Respondent marital status and highest year of school completed are negatively correlated albeit weekly also. The above was realised from the correlation coefficient stated as – 0.005. This isn’t any different from past findings as was pointed out. Marital status tends to influence how far one can school. Studies by Education office in Canada found that married women find it three times harder to go on schooling than those not married (). Other responsibilities and family duties are quoted as being the strain to continued education. However findings from the study show mean highest c schooling years between married and unmarried being almost similar although further test may be required so as to tell whether the difference noticed is significant or not. Those widowed had the least mean years in school while those not married had the highest mean years in school, these findings collaborate past studies as stated above.